ABBOT, FOWLE, ALDRICHI VARIATION OF THE SUN 311 



In the year 1903 we found indications that the radiation of 

 the sun is not constant from day to day.^ It has been a main 

 object of the work to ascertain if these apparent variations of 

 the sun are really solar, or are due to some accidental or atmos- 

 pheric influences not fully eliminated. As early as the year 1910 

 it had been shown that practically equal solar-constant values 

 were obtained on good days at sea-level, at 1730 and at 4420 

 meters elevation, and it had been shown that the apparent fluc- 

 tuations of the solar radiation found on Mount Wilson from day 

 to day marched by regular steps from high to low values and 

 return, not fluctuating wildly as they would have done had they 

 been due to experimental error. Accordingly it seemed from the 

 first consideration (namely that altitude did not appear to affect 

 the results) that the atmosphere was not the cause of the fluctu- 

 ation; and from the second consideration (namely, that the values 

 marched step by step from high to low or vice versa) that it was 

 not an accidental fluctuation. Hence, the most probable con- 

 clusion was either that the radiation of the sun is actually vari- 

 able, or that some meteoric or other matter, by interposition 

 between the earth and the sun, alters the quantity of the radiation 

 received at the earth from day to day. The fluctuations appeared 

 to be of irregular magnitude and period, often ranging through 

 5 per cent or more, in an interval of seven or ten days. 



However probable the result just stated might appear, it could 

 not be fully verified without carrying out the observation simul- 

 taneously at two stations widely separated on the earth's surface, 

 so that no local atmospheric influence could be supposed to affect 

 both stations at once. This extension of the work was made 

 possible by the Algerian expeditions of 1911 and 1912. Solar- 

 constant determinations were made nearly simultaneously at 

 Mount Wilson, California, and Bassour, Algeria, separated by 

 about one-third of the circumference of the earth. A difference 

 of time of about eight hours generally occurred between the 

 observations, but inasmuch as the apparent fluctuations of the 

 sun seldom reach 1 per cent in a day, this difference of eight 

 hours seems not much prejudicial to the comparison, 



3 See Astrophysical Journal 19: 305. 1903. 



